Line probe assays for drug-resistant tuberculosis detectionInterpretation and reporting guide for laboratory staff and clinicians

Line probe assays for drug-resistant tuberculosis detection

Interpretation and reporting guide for laboratory staff and clinicians

iii

Contents

Abbreviations v

Glossary vi

Acknowledgements vii

About this guide viii

Introduction 1

Principle of Line Probe Assay 2

GenoType MTBDRplus Version 2 3

GenoType MTBDRsl Version 2 4

Line Probe Assay interpretation and reporting 4

Revisions to manufacturer interpretations 5

Additional follow-up diagnostic actions to guide appropriate TB treatment 8

First Line (FL) – Line Probe Assay (LPA) interpretation 10

Rifampicin 10

Isoniazid 12

Second Line (SL) – Line Probe Assay (LPA) interpretation 14

Fluoroquinolones 14

Second line Injectables 18

Case studies: Examples of drug-resistant TB assessments based on SL-LPA 20

References 27

Annex 1. FL-LPA Reporting format – Practical examples 31

Annex 2. SL-LPA Reporting format – Practical examples 32

Annex 3. Specific nucleotide changes detected by MUT probes 34

CONTENT

v

Abbreviations

LIST OF ABBREVIATIONS

Am amikacin

CB clinical breakpoint

CC critical concentration

Cm capreomycin

DST drug-susceptibility testing

Eto ethionamide

FL-LPA first line – line probe assay

FQ fluoroquinolone

H isoniazid

Km kanamycin

Lfx levofloxacin

LPA line probe assay

NTM non-tuberculous mycobacteria

MIC minimum inhibitory concentration

MDR-TB multidrug-resistant tuberculosis

Mfx moxifloxacin

MTBC Mycobacterium tuberculosis complex

MUT mutation probe

Pto prothionamide

QRDR quinolone-resistance determining region

R resistant

RRDR Rifampicin- resistance determining region

Rif rifampicin

S susceptible

SLI second-line injectable (drug) (i.e. kanamycin, amikacin, capreomycin)

SL-LPA second line – line probe assay

TB tuberculosis

WT wild-type

vi

FIRST AND SECOND LINE DRUGS LINE PROBE ASSAYS

Glossary

Critical concentration (CC): The lowest concentration of an anti-TB agent in vitro that will inhibit the growth of 99% of phenotypically wild type strains of Mycobacterium tuberculosis complex (MTBC).

Clinical breakpoint (CB): concentration/s of an antimicrobial agent which defines a minimum inhibitory concentration (MIC) above the critical concentration that separates strains that will likely respond to treatment from those which will likely not respond to treatment. This concentration is determined by multiple variables, including in vivo ones, such as the correlation with available clinical outcome data and PK/PD data including drug dose, as well as MIC distributions and genetic markers. A dose increase can be used to overcome resistance observed at lower dosing, up until the maximum tolerated dose, and therefore a higher clinical breakpoint above which the particular drug is not recommended for use. The clinical breakpoint is used to guide individual clinical decisions in patient treatment.

Minimum inhibitory concentration (MIC): The lowest concentration of an antimicrobial agent that prevents growth of more than 99% of a microorganism in a solid medium or broth dilution susceptibility test.

Typically, when MICs that are tested using a standardised method are aggregated for one species, a single Gaussian-shaped MIC distribution is formed, which corresponds to the phenotypically wild-type (pWT) distribution for that species (i.e. the distribution for organisms that lack phenotypically detectable resistance mechanisms). Additional distributions with higher overall MICs are sometimes identified that correspond to intrinsically or naturally resistant organisms (i.e. phenotypically non wild-type distribution).

vii

Acknowledgements

GLOSSARY OF TERMS

This guide was developed as a product of the Global Laboratory Initiative (GLI) Core Group. Development was led and completed by Elisa Tagliani and Daniela Cirillo (San Raffaele Scientific Institute), with contributions from Elisa Ardizzoni, Bouke de Jong and Leen Rigouts (Institute of Tropical Medicine Antwerp). Special thanks to GLI Core Group members for their extensive contributions, including Heather Alexander, Olajumoke Tubi Abiola, Maka Akhalaia, Heidi Albert, Uladzimir Antonenka, Martina Casenghi, Petra de Haas, Kathleen England, Lucilaine Ferrazoli, Marguerite Massinga Loembe, Alaine Umubyeyi Nyaruhirira, Daniel Orozco, Kaiser Shen, Thomas Shinnick, Alena Skrahina, Sabira Tahseen, and Hung Van Nguyen. Coordination efforts and substantial technical input were provided by Dennis Falzon, Christopher Gilpin, Lice González-Angulo, Alexei Korobitsyn, Fuad Mirzayev and Karin Weyer from the World Health Organization (WHO) Global TB Programme during the finalization of this guide.

We appreciate the feedback provided by the following partners and stakeholders: Ignacio Monedero-Recuero from the Global Drug-Resistant TB Initiative; Paolo Miotto (San Raffaele Scientific Institute); Claudio Köser (University of Cambridge); Natalia Shubladze (ELI Core Group Member); and Soudeh Ehsani from the WHO Regional Office for Europe.

All reasonable precautions have been taken by the authors to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall the authors be liable for damages arising from its use.

GLI is a Working Group of the Stop TB Partnership. Development and publication of this document were made possible with financial support from the United States Agency for International Development.

Cover image by © Alicephotography. Layout and design of this guide were provided by minimum graphics.

viii

FIRST AND SECOND LINE DRUGS LINE PROBE ASSAYS

About this guide

This document has been developed to provide practical guidance for the interpretation of the most commonly used first line (FL) and second line (SL) line probe assays (LPA) (i.e. GenoType MTBDRplus V2.0 and GenoType MTBDRsl V2.0 assay; Hain Lifescience, Germany). In addition to the interpretation of the mutations identified by the two assays, this guide includes practical recommendations on additional follow up diagnostic actions that should be performed in the presence of specific mutations and the clinical implications that these mutations have on the selection of appropriate tuberculosis (TB) treatment regimens.

This guide is intended to provide information for both laboratory staff and clinicians on:

— the association of the specific mutations detected by the test with phenotypic drug resistance;

— the instances when specific resistance conferring mutations are not identified and resistance can only be inferred;

— the clinical implications of the specific mutations detected by the assays.

In addition, the guide provides specific support for staff at national and regional TB reference laboratories to better understand and manage possible discrepancies between phenotypic and genotypic drug susceptibility testing (DST).

This guide outlines the mutations identified by both FL- and SL-LPA test strips, including information related to their association with phenotypic drug resistance based on the work published by Miotto et al. (1) and with the minimum inhibitory concentration (MIC) for first and second line drugs as recently reported by WHO (2, 3). Test interpretation, the follow-up diagnostic testing, and clinical implications related to the presence of specific mutations as well of the clinical implications for instances when resistance is inferred are also reported.

Lastly, the guide illustrates individual case studies to demonstrate how information should be reported to clinicians, examples of LPA results, and a recommended reporting template.

1

INTRODUCTION

Introduction

In the past two decades, improved understanding of the molecular basis of drug resistance has led to the development of a plethora of genotypic-based methods for the rapid determination of susceptibility and resistance to anti-TB drugs. Besides the rapidity of diagnosis, molecular testing has additional advantages such as the possibility to be applied directly to clinical specimens (without the need for isolating the strain first on solid or liquid culture) and to specimens containing non-viable bacteria (e.g. bacteria killed by heat or chemical inactivation), the potential for high throughput and the decrease of laboratory biosafety requirements.

In 2008, the World Health Organization (WHO) endorsed the use of the first line (FL) line probe assay (LPA), the GenoType MTBDRplus (referred to as GenoType MTBDRplus V1), for the rapid detection of multidrug-resistant TB (MDR-TB) (4). Subsequently, newer versions of the LPA technology have been developed since 2011, including (i) the GenoType MTBDRplus version 2 (referred to as GenoType MTBDRplus V2), and (ii) the Nipro NTM+MDRTB detection kit 2 (referred to as “Nipro”, Tokyo, Japan). These newer LPAs aim to improve the sensitivity of (Mycobacterium tuberculosis complex) MTBC detection and to simultaneously detect resistance to rifampicin (Rif) and isoniazid (H).

In 2015, FIND (the Foundation for Innovative New Diagnostics) evaluated the Nipro and the GenoType MTBDRplus V2 LPAs and compared them with the GenoType MTBDRplus V1. The study demonstrated equivalence between these three commercially available LPAs for detecting TB and resistance to Rif and H (5).

The first commercial LPA for detection of resistance to second-line TB drugs was the GenoType MTBDRsl version 1.0, developed by Hain Lifescience a decade ago (referred to as GenoType MTBDRsl V1). An updated version of this assay (GenoType MTBDRsl V2) which detects the mutations associated with fluoroquinolone and second-line injectable (SLI) drug resistance detected by version 1.0, as well as additional mutations (described in the section below), became available in 2015.

The following year, WHO released recommendations for the use of the commercially available FL-LPAs (i.e. GenoType MTBDRplus V1, GenoType MTBDRplus V2, and Nipro) as the initial test instead of phenotypic drug-susceptibility testing (DST) to detect resistance to Rif and H (6). As well, recommendations for the use of GenoType MTBDRsl (V1 and V2) to detect resistance to fluoroquinolones and SLI drugs in patients with rifampicin-resistant/MDR-TB and to guide the initiation of appropriate MDR-TB treatment regimen were also issued by WHO in 2016 (7).

For a more detailed description of the placement of FL- and SL-LPA within the TB diagnostic algorithms, refer to the Global Laboratory Initiative (GLI) document GLI model TB diagnostic algorithms (updated in June 2018) (8).

The purpose of this document, which focuses on the two most widely used LPAs (i.e. GenoType MTBDRplus V2 and GenoType MTBDRsl V2), is to provide guidance to

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

2

laboratory staff and clinicians on the interpretation of the results for both, FL- and SL-LPA, and to better understand and manage possible discrepancies between phenotypic and genotypic DST.

Principle of Line Probe AssayLine probe assays are a family of DNA strip-based tests that determine the drug resistance profile of a MTBC strain through the pattern of binding of amplicons (DNA amplification products) to probes targeting the most common resistance associated mutations to first- and second-line agents and to probes targeting the corresponding wild-type (WT) DNA sequence.

LPAs are WHO-approved tests for rapid detection of drug resistance to first- and second-line agents. They can be used for testing of culture isolates (indirect testing), as well as direct testing of acid fast bacilli (AFB) smear microscopy positive specimens (FL-LPA), and both smear positive and smear negative sputum specimens (SL-LPA) (6, 7).

Mutations are detected by: (i) the binding of amplicons to probes targeting the most commonly occurring mutations (MUT probes) or (ii) inferred by the lack of hybridization (i.e. lack of binding) of the amplicons to the corresponding WT probes.

The post-hybridization reaction leads to the development of coloured bands on the test strip detecting probe binding.

It is important to remember that similarly to other rapid tests currently endorsed by WHO, LPAs have some limitations:

• Although LPA can detect the mutations that are most frequently identified in resistant strains, some mutations that confer resistance are outside the regions covered by the test and therefore resistance cannot be completely excluded even in the presence of all WT probes. Thus, in some cases additional phenotypic DST may be necessary to provide full assessment.

• Some mutations are identified specifically by MUT probes, whereas others are only inferred by the absence of binding of the amplicons to WT probes. The lack of binding of a WT probe without simultaneous binding of a mutant probe is likely caused by the presence of a resistance mutation. However, systematic errors are possible due to synonymous and non-synonymous mutations (e.g. phylogenetic mutations). Globally this is rare (<1% of isolates), but these isolates can be frequent locally (9).

• LPA is less efficient than conventional culture-based DST in finding resistance in samples harboring both drug-susceptible and -resistant bacteria (i.e. hetero-resistance). More specifically, with LPA, it is possible to detect resistant bacteria with mutations detected by the MUT probes if resistant bacteria represent at least 5% of the total population. However, resistant bacteria with mutations inferred by the absence of WT probes would probably be missed if the resistant population is less than 95% of the total bacterial population (10, 11).

Overall, sensitivities and specificities for the different drugs targeted by LPAs are reported in detail elsewhere (12, 13). Briefly, FL-LPA showed a sensitivity and specificity for the detection of Rif resistance of 96.7% and 98.8%, respectively, and for the

3

detection of H resistance, a sensitivity and specificity of 90.2% and 99.2%, respectively (12). SL-LPA (GenoType MTBDRsl V1) showed a pooled sensitivity and specificity for the detection of fluoroquinolone resistance by direct testing of 86.2% and 98.6%, respectively, and a pooled sensitivity and specificity for the detection of second-line injectables drugs resistance of 87.0% and 99.5%, respectively (13).

GenoType MTBDRplus Version 2The GenoType MTBDRplus (Figure 1a) targets specific mutations in the Rif-resistance determining region (RRDR) of the rpoB gene (from codon 505 to 533) (Figure 2) to detect Rif resistance and mutations in the inhA promoter (from -16 to -8 nucleotides upstream) and the katG (codon 315) regions to identify H resistance. The specific nu-cleotide changes detected by the test are reported in the Annex Section.

INTRODUCTION

Figure 1. Configuration of GenoType MTBDRplus V2 (a) and GenoType MTBDRsl V2 (b) strips

a (14) b (15)

Line 1 ........ Conjugate Control 2 ........ Amplification Control 3 ........ M. tuberculosis complex TUB

4 ........ rpoB Locus Control rpoB 5 ........ rpoB wild type probe 1 rpoB WT1 6 ........ rpoB wild type probe 2 rpoB WT2 7 ........ rpoB wild type probe 3 rpoB WT3 8 ........ rpoB wild type probe 4 rpoB WT4 9 ........ rpoB wild type probe 5 rpoB WT5 10 ........ rpoB wild type probe 6 rpoB WT6 11 ........ rpoB wild type probe 7 rpoB WT7 12 ........ rpoB wild type probe 8 rpoB WT8 13 ........ rpoB mutation probe 1 rpoB MUT1 14 ........ rpoB mutation probe 2A rpoB MUT2A 15 ........ rpoB mutation probe 2B rpoB MUT2B 16 ........ rpoB mutation probe 3 rpoB MUT3

17 ........ katG Locus Control katG 18 ........ katG wild type probe katG WT 19 ........ katG mutation probe 1 katG MUT1 20 ........ katG mutation probe 2 katG MUT2

21 ........ inhA Locus Control inhA 22 ........ inhA wild type probe 1 inhA WT1 23 ........ inhA wild type probe 2 inhA WT2 24 ........ inhA mutation probe 1 inhA MUT1 25 ........ inhA mutation probe 2 inhA MUT2 26 ........ inhA mutation probe 3A inhA MUT3A 27 ........ inhA mutation probe 3B inhA MUT3B

Colored marker

Line 1 ........ Conjugate Control 2 ........ Amplification Control 3 ........ M. tuberculosis complex TUB

4 ........ gyrA Locus Control gyrA 5 ........ gyrA wild type probe 1 gyrA WT1 6 ........ gyrA wild type probe 2 gyrA WT2 7 ........ gyrA wild type probe 3 gyrA WT3 8 ........ gyrA mutation probe 1 gyrA MUT1 9 ........ gyrA mutation probe 2 gyrA MUT2 10 ........ gyrA mutation probe 3A gyrA MUT3A 11 ........ gyrA mutation probe 3B gyrA MUT3B 12 ........ gyrA mutation probe 3C gyrA MUT3C 13 ........ gyrA mutation probe 3D gyrA MUT3D

14 ........ gyrB Locus Control gyrB 15 ........ gyrB wild type probe gyrB WT 16 ........ gyrB mutation probe 1 gyrB MUT1 17 ........ gyrB mutation probe 2 gyrB MUT2

18 ........ rrs Locus Control rrs 19 ........ rrs wild type probe 1 rrs WT1 20 ........ rrs wild type probe 2 rrs WT2 21 ........ rrs mutation probe 1 rrs MUT1 22 ........ rrs mutation probe 2 rrs MUT2

23 ........ eis Locus Control eis 24 ........ eis wild type probe 1 eis WT1 25 ........ eis wild type probe 2 eis WT2 26 ........ eis wild type probe 3 eis WT3 27 ........ eis mutation probe 1 eis MUT1

Colored marker

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

4

GenoType MTBDRsl Version 2The second version of GenoType MTBDRsl (Figure 1b) includes the quinolone-resistance determining region (QRDR) of gyrA (from codon 85 to 96) (Figure 3) and of gyrB (from codon 536 to 541) (16) genes for detection of resistance to fluoroquinolones and the rrs (nucleic acid position 1401, 1402 and 1484) and the eis promoter region (from -37 to -2 nucleotides upstream) for detection of resistance to SLI drugs. It is worth noting that the precise regions covered by all MUT probes have not been disclosed and that only some of the regions covered by WT probes are known (see Figures 2 and 3). The specific nucleotide changes detected by the MUT probes are reported in Annex Section.

Line Probe Assay interpretation and reportingThe LPA has two internal controls on the strip: the Conjugate Control (line 1), and the Amplification Control (line 2) (Figure 1). The Conjugate Control line should always be visible to document the efficiency of conjugate binding and substrate reaction. The Amplification Control serves as reference for the interpretation of WT and MUT probes: only those bands whose intensities are about as strong as or stronger than that of the Amplification Control band are to be considered. In case of a positive test result, the signal of the Amplification Control zone can be weak or even vanish totally. This can occur more frequently upon indirect testing, whereas it is rare upon direct testing. The absence of Amplification Control might be due to competition of the single reactions during amplification. In this case the test has been carried out correctly and can be interpreted. In case of a negative test result, both Conjugate Control and Amplification Control bands should always be visible (i.e. valid negative result). The absence of Amplification Control in a negative test indicates mistakes during setup and/or performance of the amplification reaction, or presence of amplification inhibitors. In this case, the test result is invalid and must be repeated.

The TUB reaction band (line 3) is only present if the DNA amplified is from members of the MTBC. The presence of non-tuberculous mycobacteria (NTM) in the specimen can result in random banding patterns, with several species testing positive at some rpoB WT bands due to the gene similarities among the species. Therefore, in the presence of NTM rather that MTBC bacteria, the TUB band will always be absent, and the result reported as MTBC not detected.

The gene locus control bands for the different target regions analyzed on the DNA strip are located just before their respective WT and MUT bands. These locus control bands must always be present for the assay to be considered valid for the corresponding target. However, when only one gene locus control band is missing, the results for the other genes for which the gene locus control band is present can be interpreted.

The LPA is defined as indeterminate for a specific drug or group of drugs if the corresponding locus control for that specific drug or group of drugs is missing while the test is valid (i.e. Conjugate Control and TUB bands are visible with or without the Amplification Control). In this case, the assay should be repeated before reporting the results. If the same result is obtained upon re-testing, report the results for the interpretable loci following the guideline, while report as indeterminate the result for the drugs or group of drugs for which the locus control is missing. Systematic reasons

5

INTRODUCTION

for these indeterminate results could be mutations or deletions in the locus control region, as well as the complete or partial deletion of a target gene. In this cases, sequencing should be requested to identify the specific mutation.

The WT reactions zones comprise regions of the genome with known resistance mutations. The MUT reaction zones correspond to probes that identify the most common resistance mutations of the gene interrogated.

Resistance is detected when MUT probes are developed, whereas in the absence of WT probes, resistance can only be inferred (see below for details).

The concomitant detection of all WT probes and one of the MUT probes in the cor-responding target region, indicates the presence of heteroresistance (i.e. coexistence of susceptible and resistant bacteria in the same sample). In this case the result should be reported as resistant.

Revisions to manufacturer interpretations (14, 15)Use of the term “Resistance not detected” instead of “Susceptible” to define the bacteria resistance profile

Given the limitations of LPA and in particular the fact that resistance cannot be completely excluded even in the presence of all WT probes (i.e. not all mutations that confer resistance are covered by these tests or mutations that are covered may be below the limit of detection), it is more appropriate to report the result as “Resistance detected” or “Resistance not detected”.

Differentiation of resistance into “Resistance inferred” and “Resistance detected”

The term “Resistance inferred” is used whenever one or more WT probes in regions of the gene known to confer resistance to the drug are not developed, and none of the MUT probes in the corresponding region is developed. In this case the precise mutation cannot be reported, only the region where the mutation lies is identified.

The term “Resistance detected” is used whenever one or more MUT probes identifying specific mutations conferring resistance to the drugs are developed (regardless of whether WT probes are developed or not).

Stratification of resistance mutations for Isoniazid (H) and Moxifloxacin (Mfx) into mutations associated with a “low-level increase in MIC” and “high-level increase in MIC”

Mutations conferring resistance to H and Mfx are stratified into mutations associated with a low-level increase in MIC and high-level increase in MIC depending on their associated MICs distribution. This stratification has important implications for the inclusion of H and Mfx in the treatment regimen since resistance due to mutations associated with low-level increases in MICs for H or Mfx may be overcome by increasing the drug dose.

For H, in vitro evidence suggests that when specific inhA promoter mutations, which are generally associated with low-level increase in MIC, are detected (and in the absence of any katG mutation), increasing the drug dose may be effective; thus, using H at a maximum dose of 15mg/kg per day could be considered. In the case of katG mutations, which are more commonly associated with high-level increase in MIC, the use of isoniazid even at higher dose is less likely to be effective. The presence

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

6

of combined mutations in the inhA promoter and katG gene results in substantial increases in the MIC, unlikely to be compensated for by increased dosing (17).

For Mfx, if mutations associated with MIC increases above the critical concentration (CC) but below the clinical breakpoint (CB), which are defined as mutations associated with a low-level increase in MIC, high dose Mfx (up to 800 mg daily to adults), is likely to be effective (18–21). When resistance to Mfx is inferred (i.e. the specific mutation is unknown), the presence of mutations associated with at least low-level increase in MIC is inferred and therefore a high drug dose may still be effective. However, in this case, it is recommended to perform DST for Mfx at the CB, and if available, sequencing to determine the specific mutation. If the MTBC strain is resistant to Mfx at the CB due to the presence of mutations associated with high-level increase in MIC, the drug cannot be considered as an effective medicine.

Whenever there is more than one probe per drug that provides information (e.g. concomitant detection of resistance mutations associated with different increases in MIC), the criteria used for the interpretation is that the mutations associated with high-level increase in MIC overrule mutations associated with low-level increase in MIC.

In summary, the results should be reported according the following hierarchy (where the “>” sign means overrule):

• For H: Mutation associated with high-level increase in MIC detected > Mutation associated with high-level increase in MIC inferred > Mutation associated with at least low-level increase in MIC detected > Mutation associated with at least low-level increase in MIC inferred > Resistance not detected.

• For Mfx: Mutation associated with high-level increase in MIC detected > Mutation associated with at least low-level increase in MIC detected > Mutation associated with at least low-level increase in MIC inferred > Resistance not detected.

• For Rif, levofloxacin (Lfx), amikacin (Am), kanamycin (Km) and capreomycin (Cm): Resistance detected > Resistance inferred > Resistance not detected.

7

In Summary

Case WT reaction zones MUT reaction zones Interpretation

1 All WT probes are developed

All MUT probes are not developed

Resistance not detected.

2 One or more WT probes are not developed

One or more MUT probes in the corresponding region are developed

Depending on the specific drug:– Resistance detected (Rif, SLI drugs);– Mutations associated with high-level

increase in MIC detected (H and Mfx);– Mutations associated with at least

low-level increase in MIC detected (H and Mfx).

3 One or more WT probes are not developed

No MUT probes developed

Depending on the specific drug:– Resistance inferred (Rif, SLI drugs);– Mutations associated with high-level

increase in MIC inferred (H and Mfx);– Mutations associated with at least

low-level increase in MIC inferred (H and Mfx).

4 All WT probes are developed

One MUT probe developed

Resistance detected (due to heteroresistance); Interpret according to case 2.

Exclusion of eis WT3 probe

To date there is no clear evidence that the mutation c-2a in the eis promoter region represents, on its own, a valid marker of resistance (1). Therefore, if eis WT3 probe is not developed, the test interpretation for Km has been revised into “Resistance not detected”.

Interpretation of resistance profile for ethionamide and prothionamide

Mutations leading to an overexpression of inhA gene, such those detected by FL-LPA, are associated with cross-resistance to ethionamide (Eto) and prothionamide (Pto) (1, 22, 23). Therefore, if these mutations are detected by the test, resistance to Eto and Pto should be reported and the medicines not used in the treatment regimen. However, it is important to note that even in the absence of mutations in the inhA promoter region, resistance to Eto/Pto cannot be excluded. Mutations conferring resistance to Eto/Pto may in fact be present in genomic regions not targeted by LPA (e.g. ethA, ethR).

Reporting of results for kanamycin and capreomycin

WHO has recently released a rapid communication on key changes to treatment of MDR and Rif-resistant TB (24) anticipating key aspects of the new WHO MDR-TB treatment guidelines that will be released in late 2018. These changes are based on the results of a meta-analysis aimed to estimate the association of treatment success and death with the use of individual drugs, and the optimal number and duration of treatment with those drugs in patients with MDR-TB (25).

On the use of SLI drugs, the results of this meta-analysis show that compared with regimens without any injectable drug, amikacin provided modest benefits, while kanamycin and capreomycin were associated with no benefits or worse outcomes. Therefore, WHO is no longer recommending the use of kanamycin and capreomycin, due to the increased risk of treatment failure and relapse associated with their use

INTRODUCTION

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

8

in longer MDR-TB regimens. In addition, programmes using the standardized shorter MDR-TB regimen should replace kanamycin with amikacin, without waiting for existing stocks of kanamycin to be used up (24).

Acknowledging the fact that it would not be immediately possible to achieve the new WHO standards of care in every individual MDR-TB patient, and that kanamycin and capreomycin will likely be used during the transition phase, the interpretation of SL-LPA for kanamycin and capreomycin has been included in this document with the purpose to provide interim guidance.

Additional follow-up diagnostic actions to guide appropriate TB treatmentDepending on the specific region interrogated by FL- and SL-LPA, one or more follow-up diagnostic actions are either recommended or suggested as optional to better guide the choice of the treatment regimen. The decision to perform the optional follow-up diagnostic actions should be guided by considerations on the individual patient’s risk for resistance and by the prevalence of resistance in the specific geographical setting, as these factors affect the positive predictive value of the test.

Different follow-up diagnostic actions are recommended or suggested as optional, depending on the specific drug considered. These actions are briefly summarized below:

Rifampicin (Rif):

• If resistance is inferred by the absence of binding of the amplicons to WT probes (i.e. one or more WT probes not developed), sequencing of rpoB gene is suggested as optional to identify the specific mutation. For the interpretation of rpoB mutations refer to the WHO Technical guide on the use of next-generation sequencing technology (3). It is important to note that phenotypic DST performed in MGIT should not be considered as the ideal confirmatory test given that it misses well-established Rif resistance associated mutations (i.e. “disputed” mutations) (26–29).

Isoniazid (H):

• If resistance is inferred by the absence of binding of the amplicons to WT probes in the katG region (i.e. one or more WT probes not developed), sequencing of katG gene is suggested as optional to identify the specific mutation. For the interpretation of katG mutations refer to the WHO Technical guide on the use of next-generation sequencing technology (3).

• If mutations associated with low-level increase in MIC are detected (i.e. MUT probes developed in the inhA promoter region in the absence of mutations in the katG target region), sequencing of inhA coding region and katG gene is suggested as optional. This is due to the fact that the concomitant presence of additional mutations in the inhA coding region or in positions other than 315 in the katG gene (mutations not detected by GenoType MTBDRplus) (30, 31), which are globally rare but could be more frequent in specific geographical settings, may cause substantial MIC increases, too high to be compensated for by increased drug dosing.

• If mutations associated with low-level increase in MIC are inferred by the absence of binding of the amplicons to WT probes in the inhA promoter region (and no mutations

9

in the katG target region are detected), it is recommended to repeat the testing to confirm the result. Optional follow-up diagnostic actions include sequencing of inhA promoter to identify the specific mutation or perform phenotypic DST for H.

Moxifloxacin (Mfx):

• If mutations associated with low-level increase in MIC are detected (i.e. MUT1, MUT2, MUT3A probes developed in gyrA and/or MUT1, MUT2 probes developed in gyrB regions), phenotypic DST for Mfx is recommended to exclude resistance at CB.

• If mutations associated with low-level increase in MIC are inferred by the absence of binding of the amplicons to WT probes in the gyrA or gyrB regions (i.e. WT probes not developed), phenotypic DST for Mfx is recommended to exclude resistance at CB. Optional follow-up actions include sequencing of gyrA and/or gyrB QRDR to identify the specific mutation and/or to perform phenotypic DST for Mfx (and/or Lfx) at CC (depending on laboratory capacity).

Amikacin (Am), kanamycin (Km),1 capreomycin (Cm):1

• If resistance is inferred by the absence of binding of the amplicons to WT probes in the rrs region (i.e. one or more WT probes not developed), it is recommended to repeat the test to confirm the result. If the result is confirmed, phenotypic DST for Am, Km, Cm should be performed to confirm resistance. Sequencing of rrs gene is suggested as optional to identify the specific mutation.

1 The use of Km and Cm in the treatment of rifampicin-resistant/MDR-TB is no longer recommended (24). Laboratories should continue to report results for Cm and Km until the recommendations are fully implemented.

INTRODUCTION

10

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

Firs

t Lin

e (F

L) –

Lin

e Pr

obe

Ass

ay (L

PA) i

nter

pret

atio

nRi

fam

pici

nFi

gure

2. R

ifam

pic

in r

esis

tan

ce d

eter

min

ing

regi

on

inte

rro

gate

d b

y G

enoT

ype

MTB

DR

plu

s

Rifa

mpi

cin-

resi

stan

ce d

eter

min

ing

regi

on (

RRD

R)

of r

poB

gene

, co

dons

cov

ered

by

the

WT

prob

es a

nd t

he s

peci

fic m

utat

ions

rec

ogni

zed

by t

he M

UT

prob

es in

Gen

oTyp

e M

TBD

Rpl

us V

er 2

.0-

for

E. c

oli v

s. M

TB c

odon

num

berin

g an

d am

ino

acid

nom

encl

atur

e (1

4).

Ove

rall,

the

sp

ecifi

city

of G

enoT

ype

MTB

DR

plus

Ver

2.0

for

Rif r

esis

tanc

e is

ver

y go

od, y

et w

hen

in d

oubt

on

the

valid

ity o

f a R

if re

sist

ance

res

ult,

req

uest

rp

oB s

eque

ncin

g as

gol

d st

anda

rd.

505

508

509

511

513

514

515

516

518

522

526

531

533

MU

T 1

D51

6V (

D4

35V

)M

UT

2A

H52

6Y (

H4

45Y

)M

UT

2B H

526D

(H

445

D)

MU

T 3

S53

1L (

S45

0L)

rpo

B W

T1rp

oB

WT3

rpo

B W

T5rp

oB

WT7

rpo

B W

T8

rpo

B W

T6rp

oB

WT4

rpo

B W

T2

Targ

et

reg

ion

MTB

DR

plu

s

Pro

be

Mu

tati

on

or

reg

ion

in

terr

oga

ted

Re

sult

in

terp

reta

tio

nA

dd

itio

nal

dia

gn

ost

ic

acti

on

aC

linic

al

imp

licat

ion

s

rpoB

WT1

rpoB

WT1

no

t de

velo

ped

Mut

atio

n(s

) at c

odon

s 50

5–50

9 (4

24–4

28)b

Resi

stan

ce to

rifa

mpi

cin

(Rif)

in

ferr

ed

Op

tio

nal

: Per

form

se

quen

cing

of r

poB

to

iden

tify

the

spec

ific

mut

atio

n.

Rifa

mpi

cin

is n

ot e

ffec

tive

rpoB

WT2

rpoB

WT2

no

t de

velo

ped

Mut

atio

n(s

) at c

odon

s 51

0–51

3 (4

29–4

32)b

Resi

stan

ce to

Rif

infe

rred

Op

tio

nal

: Per

form

se

quen

cing

of r

poB

to

iden

tify

the

spec

ific

mut

atio

n.

Rifa

mpi

cin

is n

ot e

ffec

tive

rpoB

W

T2/3

rpoB

WT2

and

WT3

no

t de

velo

ped

Mut

atio

n(s

) at c

odon

s 51

0–51

7 (4

29–4

36)b

Resi

stan

ce to

Rif

infe

rred

Op

tio

nal

: Per

form

se

quen

cing

of r

poB

to

iden

tify

the

spec

ific

mut

atio

n.

Rifa

mpi

cin

is n

ot e

ffec

tive

11

Targ

et

reg

ion

MTB

DR

plu

s P

rob

eM

uta

tio

n o

r re

gio

n

inte

rro

gate

dR

esu

lt

inte

rpre

tati

on

Ad

dit

ion

al d

iag

no

stic

ac

tio

na

Clin

ical

im

plic

atio

ns

rpoB

W

T3/4

rpoB

MU

T1 d

evel

oped

D51

6V (D

435V

)bRe

sist

ance

to R

if de

tect

edN

o ad

ditio

nal d

iagn

ostic

ac

tion

requ

ired.

Rifa

mpi

cin

is n

ot e

ffec

tive

rpoB

WT3

, WT4

and

MU

T1

no

t de

velo

ped

Mut

atio

n(s

) at c

odon

s 51

3–51

9 (4

32–4

38) b

Resi

stan

ce to

Rif

infe

rred

Op

tio

nal

: Per

form

se

quen

cing

of r

poB

to

iden

tify

the

spec

ific

mut

atio

n.

Rifa

mpi

cin

is n

ot e

ffec

tive

rpoB

W

T4/5

rpoB

WT4

and

WT5

no

t de

velo

ped

Mut

atio

n(s

) at c

odon

s 51

6–52

2 (4

35–4

41)b

Resi

stan

ce to

Rif

infe

rred

Op

tio

nal

: Per

form

se

quen

cing

of r

poB

to

iden

tify

the

spec

ific

mut

atio

n.

Rifa

mpi

cin

is n

ot e

ffec

tive

rpoB

W

T5/6

rpoB

WT5

and

WT6

no

t de

velo

ped

Mut

atio

n(s

) at c

odon

s 51

8–52

5 (4

37–4

44)b

Resi

stan

ce to

Rif

infe

rred

Op

tio

nal

: Per

form

se

quen

cing

of r

poB

to

iden

tify

the

spec

ific

mut

atio

n.

Rifa

mpi

cin

is n

ot e

ffec

tive

rpoB

WT7

rpoB

MU

T2A

dev

elop

edH

526Y

(H44

5Y) b

Resi

stan

ce to

Rif

dete

cted

No

addi

tiona

l dia

gnos

tic

actio

n re

quire

dRi

fam

pici

n is

not

eff

ectiv

e

rpoB

MU

T2B

dev

elop

edH

526D

(H44

5D)b

Resi

stan

ce to

Rif

dete

cted

No

addi

tiona

l dia

gnos

tic

actio

n re

quire

dRi

fam

pici

n is

not

eff

ectiv

e

rpoB

WT7

, MU

T2A

and

M

UT2

B n

ot

deve

lope

dM

utat

ion

(s) a

t cod

ons

526–

529

(445

–448

) bRe

sist

ance

to R

if in

ferr

ed

Op

tio

nal

: Per

form

se

quen

cing

of r

poB

to

iden

tify

the

spec

ific

mut

atio

n.

Rifa

mpi

cin

is n

ot e

ffec

tive

rpoB

WT8

rpoB

MU

T3 d

evel

oped

S531

L (S

450L

) bRe

sist

ance

to R

if de

tect

edN

o ad

ditio

nal d

iagn

ostic

ac

tion

requ

ired

Rifa

mpi

cin

is n

ot e

ffec

tive

rpoB

WT8

and

MU

T3 n

ot

deve

lope

dM

utat

ion(

s) a

t co

don

530–

533

(449

–452

) bRe

sist

ance

to R

if in

ferr

ed

Op

tio

nal

: Per

form

se

quen

cing

of r

poB

to

iden

tify

the

spec

ific

mut

atio

n.

Rifa

mpi

cin

is n

ot e

ffec

tive

a Th

e de

cisi

on t

o pe

rfor

m t

he o

ptio

nal d

iagn

ostic

act

ions

sho

uld

be g

uide

d by

con

side

ratio

ns o

n th

e in

divi

dual

pat

ient

’s r

isk

grou

p fo

r re

sist

ance

and

by

the

prev

alen

ce o

f res

ista

nce

in t

he s

peci

fic

geog

raph

ical

set

ting,

as

thes

e fa

ctor

s af

fect

the

pos

itive

pre

dict

ive

valu

e of

the

tes

t. S

ilent

mut

atio

ns m

ay b

e of

gre

ater

con

cern

in lo

w r

ifam

pici

n-re

sist

ant p

reva

lenc

e se

ttin

gs.

b

MTB

cod

on n

umbe

ring

acc

ordi

ng t

o A

ndre

et a

l (32

) is

repo

rted

in p

aren

thes

is.

FIRST LINE (FL) – LINE PROBE ASSAY (LPA) INTERPRETATION

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

12

Ison

iazi

d

Targ

et

reg

ion

MTB

DR

plu

s P

rob

eM

uta

tio

n o

r re

gio

n

inte

rro

gate

dR

esu

lt

inte

rpre

tati

on

Ad

dit

ion

al d

iag

no

stic

ac

tio

na

Clin

ical

im

plic

atio

ns

katG

WT

katG

MU

T1 o

r M

UT2

de

velo

ped

S315

T1 /

S315

T2M

utat

ion

asso

ciat

ed w

ith

high

-leve

l inc

reas

e in

MIC

de

tect

ed.

No

addi

tiona

l dia

gnos

tic

actio

n re

quire

d

Ison

iazi

d is

unl

ikel

y to

be

effe

ctiv

e ev

en a

t hig

h do

se

(17)

.

katG

WT,

MU

T1 a

nd M

UT2

n

ot

deve

lope

d b

Mut

atio

n(s

) at c

odon

315

Mut

atio

n as

soci

ated

with

hi

gh-le

vel i

ncre

ase

in M

IC

infe

rred

.

Op

tio

nal

: Per

form

se

quen

cing

of k

atG

to

iden

tify

the

spec

ific

mut

atio

n.

Ison

iazi

d is

unl

ikel

y to

be

effe

ctiv

e ev

en a

t hig

h do

se

(17)

.

inhA

WT1

inhA

MU

T1 d

evel

oped

c-15

t

Mut

atio

n as

soci

ated

with

at

leas

t low

-leve

l inc

reas

e in

M

IC d

etec

ted.

Resi

stan

ce to

Eto

/Pto

de

tect

ed.

Op

tio

nal

c : P

erfo

rm

sequ

enci

ng o

f inh

A c

odin

g re

gion

and

kat

G g

ene.

No

addi

tiona

l dia

gnos

tic

actio

n fo

r Et

o/P

to.

Ison

iazi

d at

hig

h do

se is

lik

ely

effe

ctiv

e (1

7) .

Ethi

onam

ide/

prot

hion

amid

e ar

e no

t eff

ectiv

e.

inhA

MU

T2 d

evel

oped

a-16

g d

Mut

atio

n lik

ely

asso

ciat

ed

with

at l

east

low

-leve

l in

crea

se in

MIC

det

ecte

d.Re

sist

ance

to E

to/P

to li

kely

de

tect

ed.

Op

tio

nal

c : P

erfo

rm

sequ

enci

ng o

f inh

A c

odin

g re

gion

and

kat

G g

ene.

No

addi

tiona

l dia

gnos

tic

actio

n fo

r Et

o/P

to.

Ison

iazi

d at

hig

h do

se is

lik

ely

effe

ctiv

e (2

3).

Ethi

onam

ide/

prot

hion

amid

e ar

e lik

ely

not e

ffec

tive.

InhA

WT1

, MU

T1 a

nd M

UT2

n

ot

deve

lope

dM

utat

ion(s)

in th

e -1

5 re

gion

d

Mut

atio

n lik

ely

asso

ciat

ed

with

at l

east

low

-leve

l in

crea

se in

MIC

infe

rred

.Re

sist

ance

to E

to/P

to li

kely

in

ferr

ed.

Rec

om

men

de

d: R

epea

t SL

-LPA

to c

onfir

m th

e re

sult.

Op

tio

nal

: –

Perf

orm

seq

uenc

ing

to

iden

tify

spec

ific

mut

atio

n.–

Perf

orm

phe

noty

pic

DST

fo

r H

, Eto

/Pto

.

Ison

iazi

d at

hig

h do

se is

lik

ely

effe

ctiv

e (2

3).

Ethi

onam

ide/

prot

hion

amid

e ar

e lik

ely

not e

ffec

tive.

13

Targ

et

reg

ion

MTB

DR

plu

s P

rob

eM

uta

tio

n o

r re

gio

n

inte

rro

gate

dR

esu

lt

inte

rpre

tati

on

Ad

dit

ion

al d

iag

no

stic

ac

tio

na

Clin

ical

im

plic

atio

ns

inhA

WT2

InhA

MU

T3A

dev

elop

edt-

8c d

Mut

atio

n lik

ely

asso

ciat

ed

with

at l

east

low

-leve

l in

crea

se in

MIC

det

ecte

d.Re

sist

ance

to E

to/P

to li

kely

de

tect

ed.

Op

tio

nal

c : P

erfo

rm

sequ

enci

ng o

f inh

A c

odin

g re

gion

and

kat

G g

ene.

No

addi

tiona

l dia

gnos

tic

actio

n fo

r Et

o/P

to.

Ison

iazi

d at

hig

h do

se is

lik

ely

effe

ctiv

e (2

3).

Ethi

onam

ide/

prot

hion

amid

e ar

e lik

ely

not e

ffec

tive.

InhA

MU

T3B

deve

lope

dt-

8a d

Mut

atio

n lik

ely

asso

ciat

ed

with

at l

east

low

-leve

l in

crea

se in

MIC

det

ecte

d.Re

sist

ance

to E

to/P

to li

kely

de

tect

ed.

Op

tio

nal

c : P

erfo

rm

sequ

enci

ng o

f inh

A c

odin

g re

gion

and

kat

G g

ene.

No

addi

tiona

l dia

gnos

tic

actio

n fo

r Et

o/P

to.

Ison

iazi

d at

hig

h do

se is

lik

ely

effe

ctiv

e (2

3).

Ethi

onam

ide/

prot

hion

amid

e ar

e lik

ely

not e

ffec

tive.

InhA

WT2

, MU

T3A

and

M

UT3

B n

ot

deve

lope

dM

utat

ion

in th

e -8

reg

ion

d

Mut

atio

n lik

ely

asso

ciat

ed

with

at l

east

low

-leve

l in

crea

se in

MIC

infe

rred

.Re

sist

ance

to E

to/P

to li

kely

in

ferr

ed.

Rec

om

men

de

d: R

epea

t FL

-LPA

to c

onfir

m th

e re

sult.

Op

tio

nal

: –

Perf

orm

seq

uenc

ing

to

iden

tify

spec

ific

mut

atio

n.–

Perf

orm

phe

noty

pic

DST

fo

r H

, Eto

/Pto

.

Ison

iazi

d at

hig

h do

se is

lik

ely

effe

ctiv

e (2

3).

Ethi

onam

ide/

prot

hion

amid

e ar

e lik

ely

not e

ffec

tive.

a Th

e de

cisi

on t

o pe

rfor

m t

he o

ptio

nal f

ollo

w-u

p di

agno

stic

act

ions

sho

uld

be g

uide

d by

con

side

ratio

ns o

n th

e in

divi

dual

pat

ient

’s r

isk

grou

p fo

r re

sist

ance

and

by

the

prev

alen

ce o

f res

ista

nce

in t

he s

peci

fic

geog

raph

ical

set

ting,

as

thes

e fa

ctor

s af

fect

the

pos

itive

pre

dict

ive

valu

e of

the

tes

t.

b Th

e pa

rtia

l or

who

le d

elet

ion

of t

he k

atG

gen

e, w

hich

is a

ssoc

iate

d w

ith h

igh-

leve

l inc

reas

e in

MIC

, res

ults

in t

he c

ompl

ete

abse

nce

of k

atG

locu

s ba

nds

(i.e.

kat

G lo

cus

cont

rol,

WT

and

MU

T pr

obes

are

not

de

velo

ped

).c

The

conc

omita

nt p

rese

nce

of a

dditi

onal

mut

atio

ns in

the

inhA

cod

ing

regi

on o

r in

pos

ition

s ot

her

than

315

in t

he k

atG

gen

e (m

utat

ions

not

det

ecte

d by

Gen

oTyp

e M

TBD

Rpl

us) (

30, 3

1), w

hich

are

glo

bally

ra

re b

ut c

ould

be

mor

e fr

eque

nt in

spe

cific

geo

grap

hica

l set

tings

, may

cau

se s

ubst

antia

l MIC

incr

ease

s, t

oo h

igh

to b

e co

mpe

nsat

ed fo

r by

incr

ease

d dr

ug d

osin

g.d

Add

ition

al d

ata

corr

elat

ing

thes

e m

utat

ions

with

phe

noty

pic

DST

for

ison

iazi

d is

nee

ded

to in

crea

se t

he c

onfid

ence

in t

he a

ssoc

iatio

n of

the

se m

utat

ions

with

dru

g re

sist

ance

.

FIRST LINE (FL) – LINE PROBE ASSAY (LPA) INTERPRETATION

14

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

Seco

nd L

ine

(SL)

– L

ine

Prob

e A

ssay

(LPA

) int

erpr

etat

ion

Fluo

roqu

inol

ones

Figu

re 3

. Qu

ino

lon

e re

sist

ance

det

erm

inin

g re

gio

n o

f g

yrA

gen

e in

terr

oga

ted

by

Gen

oTyp

e M

TBD

Rsl

.

Qui

nolo

ne-r

esis

tanc

e de

term

inin

g re

gion

(QRD

R) o

f gyr

A g

ene,

the

codo

ns c

over

ed b

y th

e W

T pr

obes

and

the

spec

ific

mut

atio

ns (b

oth

amin

o ac

id a

nd n

ucle

otid

e ch

ange

s) r

ecog

nize

d by

the

MU

T pr

obes

in G

enoT

ype

MTB

DRs

l Ver

2.0

(15)

.

8586

8788

8990

9192

9394

9596

97

MU

T1 A

90V

gcg

>g

tgM

UT

3A D

94A

gac

>gc

cM

UT

3B D

94N

gac

>aa

cM

UT

3B D

94Y

gac

>ta

cM

UT

3C D

94G

gac

>g

gcM

UT

3D D

94H

gac

>ca

c

gyr

A W

T1g

yrA

WT3

gyr

A W

T2

MU

T2 S

91P

tcg

>cc

g

Targ

et

reg

ion

MTB

DR

sl

Pro

be

Mu

tati

on

or

reg

ion

in

terr

oga

ted

Re

sult

in

terp

reta

tio

nA

dd

itio

nal

dia

gn

ost

ic

acti

on

aC

linic

al

imp

licat

ion

s

gyrA

WT

1gy

rA W

T1 n

ot

deve

lope

dC

odon

(s) 8

5–89

Resi

stan

ce to

Lfx

infe

rred

.M

utat

ion

asso

ciat

ed w

ith a

t le

ast l

ow-le

vel i

ncre

ase

in

MIC

for

Mfx

infe

rred

.

Rec

om

men

de

d: P

erfo

rm

phen

otyp

ic D

ST fo

r M

fx a

t C

B to

exc

lude

res

ista

nce.

Op

tio

nal

:–

Perf

orm

seq

uenc

ing

of

gyrA

QRD

R to

iden

tify

spec

ific

mut

atio

n;–

Perf

orm

phe

noty

pic

DST

fo

r Lf

x an

d M

fx a

t CC

.

Levo

floxa

cin

is n

ot e

ffec

tive.

Mox

iflox

acin

cou

ld b

e us

ed a

t hig

her

dose

. The

reg

imen

sho

uld

be r

e-ev

alua

ted

base

d on

phe

noty

pic

DST

re

sults

at C

B.

No

te. T

hese

rec

omm

enda

tions

do

not a

pply

if s

eque

ncin

g, a

vaila

ble

befo

re tr

eatm

ent i

nitia

tion,

iden

tifies

m

utat

ions

not

ass

ocia

ted

with

re

sist

ance

to F

Q, o

r if

phen

otyp

ic

DST

sho

ws

susc

eptib

ility

at C

C.

15

Targ

et

reg

ion

MTB

DR

sl

Pro

be

Mu

tati

on

or

reg

ion

in

terr

oga

ted

Re

sult

in

terp

reta

tio

nA

dd

itio

nal

dia

gn

ost

ic

acti

on

aC

linic

al

imp

licat

ion

s

gyrA

WT2

gyrA

MU

T1

deve

lope

dA

90V

Resi

stan

ce to

Lfx

det

ecte

d.M

utat

ion

asso

ciat

ed w

ith a

t le

ast l

ow-le

vel i

ncre

ase

in

MIC

for

Mfx

det

ecte

d.

Rec

om

men

de

d: P

erfo

rm

phen

otyp

ic D

ST fo

r M

fx a

t C

B to

exc

lude

res

ista

nce.

Levo

floxa

cin

is n

ot e

ffec

tive.

Mox

iflox

acin

cou

ld b

e us

ed a

t hig

her

dose

. The

reg

imen

sho

uld

be r

e-ev

alua

ted

base

d on

phe

noty

pic

DST

re

sults

at C

B.

gyrA

MU

T2

deve

lope

dS9

1P

Resi

stan

ce to

Lfx

det

ecte

d.M

utat

ion

asso

ciat

ed w

ith a

t le

ast l

ow-le

vel i

ncre

ase

in

MIC

for

Mfx

det

ecte

d.

Rec

om

men

de

d: P

erfo

rm

phen

otyp

ic D

ST fo

r M

fx a

t C

B to

exc

lude

res

ista

nce.

Levo

floxa

cin

is n

ot e

ffec

tive.

Mox

iflox

acin

cou

ld b

e us

ed a

t hig

her

dose

. The

reg

imen

sho

uld

be r

e-ev

alua

ted

base

d on

phe

noty

pic

DST

re

sults

at C

B.

gyrA

WT2

, MU

T1

and

MU

T2 n

ot

deve

lope

dC

odon

(s) 8

9–93

Resi

stan

ce to

Lfx

infe

rred

.M

utat

ion

asso

ciat

ed w

ith a

t le

ast l

ow-le

vel i

ncre

ase

in

MIC

for

Mfx

infe

rred

.

Rec

om

men

de

d: P

erfo

rm

phen

otyp

ic D

ST fo

r M

fx a

t C

B to

exc

lude

res

ista

nce.

Op

tio

nal

:–

Perf

orm

seq

uenc

ing

of

gyrA

QRD

R to

iden

tify

spec

ific

mut

atio

n;–

Perf

orm

phe

noty

pic

DST

fo

r Lf

x an

d M

fx a

t CC

.

Levo

floxa

cin

is n

ot e

ffec

tive.

Mox

iflox

acin

cou

ld b

e us

ed a

t hig

her

dose

. The

reg

imen

sho

uld

be r

e-ev

alua

ted

base

d on

phe

noty

pic

DST

re

sults

at C

B.

No

te. T

hese

rec

omm

enda

tions

do

not a

pply

if s

eque

ncin

g, a

vaila

ble

befo

re tr

eatm

ent i

nitia

tion,

iden

tifies

m

utat

ions

not

ass

ocia

ted

with

re

sist

ance

to F

Q, o

r if

phen

otyp

ic

DST

sho

ws

susc

eptib

ility

at C

C.

gyrA

WT3

gyrA

MU

T3A

de

velo

ped

D94

A

Resi

stan

ce to

Lfx

det

ecte

d.M

utat

ion

asso

ciat

ed w

ith a

t le

ast l

ow-le

vel i

ncre

ase

in

MIC

for

Mfx

det

ecte

d.

Rec

om

men

de

d: P

erfo

rm

phen

otyp

ic D

ST fo

r M

fx a

t C

B to

exc

lude

res

ista

nce.

Levo

floxa

cin

is n

ot e

ffec

tive.

Mox

iflox

acin

cou

ld b

e us

ed a

t hig

her

dose

. The

reg

imen

sho

uld

be r

e-ev

alua

ted

base

d on

phe

noty

pic

DST

re

sults

at C

B.

gyrA

MU

T3B

de

velo

ped

D94

N o

r D

94Y

Resi

stan

ce to

Lfx

det

ecte

d.M

utat

ion

asso

ciat

ed w

ith

high

-leve

l inc

reas

e in

MIC

fo

r M

fx d

etec

ted.

No

addi

tiona

l dia

gnos

tic

actio

n re

quire

d.Le

voflo

xaci

n is

not

eff

ectiv

e.

Mox

iflox

acin

is n

ot e

ffec

tive.

SECOND LINE (SL) – LINE PROBE ASSAY (LPA) INTERPRETATION

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

16

Targ

et

reg

ion

MTB

DR

sl

Pro

be

Mu

tati

on

or

reg

ion

in

terr

oga

ted

Re

sult

in

terp

reta

tio

nA

dd

itio

nal

dia

gn

ost

ic

acti

on

aC

linic

al

imp

licat

ion

s

gyrA

WT3

gyrA

MU

T3C

de

velo

ped

D94

G

Resi

stan

ce to

Lfx

det

ecte

d.M

utat

ion

asso

ciat

ed w

ith

high

-leve

l inc

reas

e in

MIC

fo

r M

fx d

etec

ted.

No

addi

tiona

l dia

gnos

tic

actio

n re

quire

d.Le

voflo

xaci

n is

not

eff

ectiv

e.

Mox

iflox

acin

is n

ot e

ffec

tive.

gyrA

MU

T3D

de

velo

ped

D94

HRe

sist

ance

to L

fx d

etec

ted.

Mut

atio

n as

soci

ated

with

hi

gh-le

vel i

ncre

ase

in M

IC

for

Mfx

det

ecte

d.

No

addi

tiona

l dia

gnos

tic

actio

n re

quire

d.Le

voflo

xaci

n is

not

eff

ectiv

e.

Mox

iflox

acin

is n

ot e

ffec

tive.

gyrA

WT3

, MU

T3A

, M

UT3

B, M

UT3

C,

MU

T3D

no

t de

velo

ped

Cod

on(s

) 92–

96

Resi

stan

ce to

Lfx

infe

rred

.M

utat

ion

asso

ciat

ed w

ith a

t le

ast l

ow-le

vel i

ncre

ase

in

MIC

for

Mfx

infe

rred

.

Rec

om

men

de

d: P

erfo

rm

phen

otyp

ic D

ST fo

r M

fx a

t C

B to

exc

lude

res

ista

nce.

Op

tio

nal

:–

P erf

orm

seq

uenc

ing

of

gyrA

QRD

R to

iden

tify

spec

ific

mut

atio

n;–

Perf

orm

phe

noty

pic

DST

fo

r Lf

x an

d M

fx a

t CC

.

Levo

floxa

cin

is n

ot e

ffec

tive.

Mox

iflox

acin

cou

ld b

e us

ed a

t hig

her

dose

. The

reg

imen

sho

uld

be r

e-ev

alua

ted

base

d on

phe

noty

pic

DST

re

sults

at C

B.

No

te. T

hese

rec

omm

enda

tions

do

not a

pply

if s

eque

ncin

g, a

vaila

ble

befo

re tr

eatm

ent i

nitia

tion,

iden

tifies

m

utat

ions

not

ass

ocia

ted

with

re

sist

ance

to F

Q, o

r if

phen

otyp

ic

DST

sho

ws

susc

eptib

ility

at C

C.

a Th

e de

cisi

on t

o pe

rfor

m t

he o

ptio

nal f

ollo

w-u

p di

agno

stic

act

ions

sho

uld

be g

uide

d by

con

side

ratio

ns o

n th

e in

divi

dual

pat

ient

’s r

isk

grou

p fo

r re

sist

ance

and

by

the

prev

alen

ce o

f res

ista

nce

in t

he s

peci

fic

geog

raph

ical

set

ting,

as

thes

e fa

ctor

s af

fect

the

pos

itive

pre

dict

ive

valu

e of

the

tes

t.

17

Targ

et

reg

ion

MTB

DR

sl

Pro

be

Mu

tati

on

or

reg

ion

in

terr

oga

ted

Re

sult

in

terp

reta

tio

nA

dd

itio

nal

dia

gn

ost

ic

acti

on

aC

linic

al

imp

licat

ion

s

gyrB

WT

gyrB

MU

T1

deve

lope

dN

538D

(cod

on 4

99)b

Resi

stan

ce to

Lfx

det

ecte

d.M

utat

ion

asso

ciat

ed w

ith

at le

ast l

ow-le

vel i

ncre

ase

in M

IC fo

r M

fx d

etec

ted.

Rec

om

men

de

d: P

erfo

rm

phen

otyp

ic D

ST fo

r M

fx to

ex

clud

e re

sist

ance

at C

B.

Levo

floxa

cin

is n

ot e

ffec

tive.

Mox

iflox

acin

cou

ld b

e us

ed a

t hig

her

dose

. The

reg

imen

sho

uld

be r

e-ev

alua

ted

base

d on

phe

noty

pic

DST

re

sults

at C

B.

gyrB

MU

T2

deve

lope

dE5

40V

(cod

on 5

01)b

Resi

stan

ce to

Lfx

det

ecte

d.M

utat

ion

asso

ciat

ed w

ith

at le

ast l

ow-le

vel i

ncre

ase

in M

IC fo

r M

fx d

etec

ted.

Rec

om

men

de

d: P

erfo

rm

phen

otyp

ic D

ST fo

r M

fx to

ex

clud

e re

sist

ance

at C

B.

Levo

floxa

cin

is n

ot e

ffec

tive.

Mox

iflox

acin

cou

ld b

e us

ed a

t hig

her

dose

. The

reg

imen

sho

uld

be r

e-ev

alua

ted

base

d on

phe

noty

pic

DST

re

sults

at C

B.

gyrB

WT,

MU

T1

and

MU

T2 n

ot

deve

lope

d

Cod

on(s

) 536

–541

(cod

on 4

97–5

02)b

Resi

stan

ce to

Lfx

infe

rred

.M

utat

ion

asso

ciat

ed w

ith

at le

ast l

ow-le

vel i

ncre

ase

in M

IC fo

r M

fx in

ferr

ed.

Rec

om

men

de

d: P

erfo

rm

phen

otyp

ic D

ST fo

r M

fx to

ex

clud

e re

sist

ance

at C

B.

Op

tio

nal

:–

Perf

orm

seq

uenc

ing

of

gyrA

QRD

R to

iden

tify

spec

ific

mut

atio

n;–

Perf

orm

phe

noty

pic

DST

fo

r Lf

x an

d M

fx a

t CC

.

Levo

floxa

cin

is n

ot e

ffec

tive.

Mox

iflox

acin

cou

ld b

e us

ed a

t hig

her

dose

. The

reg

imen

sho

uld

be r

e-ev

alua

ted

base

d on

phe

noty

pic

DST

re

sults

at C

B.

No

te. T

hese

rec

omm

enda

tions

do

not

appl

y if

sequ

enci

ng, a

vaila

ble

befo

re

trea

tmen

t ini

tiatio

n, id

entifi

es m

utat

ions

no

t ass

ocia

ted

with

res

ista

nce

to F

Q, o

r if

phen

otyp

ic D

ST s

how

s su

scep

tibili

ty

at C

C.

a Th

e de

cisi

on t

o pe

rfor

m t

he o

ptio

nal f

ollo

w u

p di

agno

stic

act

ions

sho

uld

be g

uide

d by

con

side

ratio

ns o

n th

e in

divi

dual

pat

ient

’s r

isk

grou

p fo

r re

sist

ance

and

by

the

prev

alen

ce o

f res

ista

nce

in t

he s

peci

fic

geog

raph

ical

set

ting,

as

thes

e fa

ctor

s af

fect

the

pos

itive

pre

dict

ive

valu

e of

the

tes

t.b

Cod

on n

umbe

ring

sys

tem

acc

ordi

ng t

o C

amus

et a

l (33

).

SECOND LINE (SL) – LINE PROBE ASSAY (LPA) INTERPRETATION

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

18

Seco

nd li

ne In

ject

able

sa

Targ

et

reg

ion

MTB

DR

sl

Pro

be

Mu

tati

on

o

r re

gio

n

inte

rro

gate

d

Re

sult

in

terp

reta

tio

n f

or

Kan

amyc

in (

Km

)

Re

sult

in

terp

reta

tio

n f

or

Am

ikac

in (

Am

)

Re

sult

in

terp

reta

tio

n f

or

Cap

reo

myc

in (

Cm

)

Ad

dit

ion

al

dia

gn

ost

ic a

ctio

nb

Clin

ical

imp

licat

ion

s

rrs

WT1

rrs

MU

T1

deve

lope

da1

401g

Resi

stan

ce to

Km

de

tect

ed.

Resi

stan

ce to

Am

de

tect

ed.

Resi

stan

ce to

Cm

de

tect

ed.

No

addi

tiona

l di

agno

stic

act

ion

requ

ired.

Am

ikac

in, k

anam

ycin

an

d ca

preo

myc

in a

re

not e

ffec

tive.

rrs

WT1

and

M

UT1

no

t de

velo

ped

Mut

atio

n(s)

in th

e 14

00 r

egio

n

Resi

stan

ce to

Km

in

ferr

ed.

Resi

stan

ce to

Am

in

ferr

ed c

Resi

stan

ce to

Cm

in

ferr

ed.

Rec

om

men

de

d:

Repe

at th

e SL

-LPA

test

and

if

the

resu

lt is

co

nfirm

ed, p

erfo

rm

phen

otyp

ic D

ST fo

r A

m, K

m, C

m.

Op

tio

nal

: Per

form

se

quen

cing

to

iden

tify

spec

ific

mut

atio

n.

Kana

myc

in a

nd

capr

eom

ycin

are

like

ly

not e

ffec

tive.

Phen

otyp

ic D

ST r

esul

t sh

ould

gui

de th

e ch

oice

to

use

Am

ikac

in in

the

trea

tmen

t reg

imen

.

rrs

WT2

rrs

MU

T2

deve

lope

dg1

484t

Resi

stan

ce to

Km

de

tect

ed.

Resi

stan

ce to

Am

de

tect

ed.

Resi

stan

ce to

Cm

de

tect

ed.

No

addi

tiona

l di

agno

stic

act

ion

requ

ired.

Am

ikac

in, k

anam

ycin

an

d ca

preo

myc

in a

re

not e

ffec

tive.

rrs

WT2

and

M

UT2

no

t de

velo

ped

Mut

atio

n in

the

1484

re

gion

Resi

stan

ce to

Km

in

ferr

ed.

Resi

stan

ce to

Am

in

ferr

ed.

Resi

stan

ce to

Cm

in

ferr

ed.

Rec

om

men

de

d:

Repe

at th

e SL

-LPA

test

and

if

the

resu

lt is

co

nfirm

ed, p

erfo

rm

phen

otyp

ic D

ST fo

r A

m, K

m a

nd C

m.

Op

tio

nal

: Per

form

se

quen

cing

to

iden

tify

spec

ific

mut

atio

n.

Am

ikac

in, k

anam

ycin

an

d ca

preo

myc

in a

re

likel

y no

t eff

ectiv

e.

19

Targ

et

reg

ion

MTB

DR

sl

Pro

be

Mu

tati

on

o

r re

gio

n

inte

rro

gate

d

Re

sult

in

terp

reta

tio

n f

or

Kan

amyc

in (

Km

)

Re

sult

in

terp

reta

tio

n f

or

Am

ikac

in (

Am

)

Re

sult

in

terp

reta

tio

n f

or

Cap

reo

myc

in (

Cm

)

Ad

dit

ion

al

dia

gn

ost

ic a

ctio

nb

Clin

ical

imp

licat

ion

s

eis

WT1

eis

WT1

not

de

velo

ped

Mut

atio

n(s)

in t

he

-37

regi

onRe

sist

ance

to K

m

infe

rred

.Re

sist

ance

to A

m

not d

etec

ted.

Resi

stan

ce to

Cm

no

t det

ecte

d.O

pti

on

al: P

erfo

rm

sequ

enci

ng to

id

entif

y sp

ecifi

c m

utat

ion

Am

ikac

in is

like

ly

effe

ctiv

e.

Cap

reom

ycin

is li

kely

ef

fect

ive.

Kana

myc

in is

not

ef

fect

ive.

eis

WT2

eis

MU

T1

deve

lope

dc-

14t

Resi

stan

ce to

Km

de

tect

ed.

Resi

stan

ce to

Am

no

t det

ecte

d.Re

sist

ance

to C

m

not d

etec

ted.

Op

tio

nal

: Per

form

se

quen

cing

to

iden

tify

spec

ific

mut

atio

n

Am

ikac

in is

like

ly

effe

ctiv

e.

Cap

reom

ycin

is li

kely

ef

fect

ive.

Kana

myc

in is

not

ef

fect

ive.

eis

WT2

and

M

UT1

not

de

velo

ped

Mut

atio

n(s)

in th

e -1

0 to

-15

reg

ion

Resi

stan

ce to

Km

in

ferr

ed.

Resi

stan

ce to

Am

no

t det

ecte

d.Re

sist

ance

to C

m

not d

etec

ted.

Op

tio

nal

: Per

form

se

quen

cing

to

iden

tify

spec

ific

mut

atio

n

Am

ikac

in is

like

ly

effe

ctiv

e.

Cap

reom

ycin

is li

kely

ef

fect

ive.

Kana

myc

in is

like

ly n

ot

effe

ctiv

e.

eis

WT3

eis

WT3

not

de

velo

ped

Mut

atio

n(s)

in th

e -2

reg

ion

Not

e. N

o ev

iden

ce

that

mut

atio

ns in

this

regi

on a

re a

ssoc

iate

d w

ith r

esis

tanc

e (1

)

Resi

stan

ce to

Km

no

t det

ecte

d.Re

sist

ance

to A

m

not d

etec

ted.

Resi

stan

ce to

Cm

no

t det

ecte

d.O

pti

on

al: P

erfo

rm

sequ

enci

ng to

id

entif

y sp

ecifi

c m

utat

ion

Am

ikac

in is

like

ly

effe

ctiv

e.

Cap

reom

ycin

is li

kely

ef

fect

ive.

Kana

myc

in is

like

ly

effe

ctiv

e.

a W

HO

is n

o lo

nger

rec

omm

endi

ng t

he u

se o

f kan

amyc

in a

nd c

apre

omyc

in d

ue t

o th

e in

crea

sed

risk

of t

reat

men

t fai

lure

and

rel

apse

ass

ocia

ted

with

the

ir us

e in

long

er M

DR-

TB r

egim

ens

(23)

. The

in

terp

reta

tion

of S

L-LP

A fo

r ka

nam

ycin

and

cap

reom

ycin

has

bee

n in

clud

ed in

thi

s do

cum

ent w

ith t

he p

urpo

se t

o pr

ovid

e in

teri

m g

uida

nce

duri

ng t

he t

rans

ition

pha

se.

b Th

e de

cisi

on t

o pe

rfor

m a

dditi

onal

dia

gnos

tic a

ctio

ns in

dica

ted

as o

ptio

nal s

houl

d be

gui

ded

by c

onsi

dera

tions

on

the

indi

vidu

al p

atie

nt’s

ris

k gr

oup

for

resi

stan

ce a

nd b

y th

e pr

eval

ence

of r

esis

tanc

e in

th

e sp

ecifi

c ge

ogra

phic

al s

ettin

g, a

s th

ese

fact

ors

affe

ct t

he p

ositi

ve p

redi

ctiv

e va

lue

of t

he t

est.

c

As

no M

UT

prob

e in

the

rrs

1400

reg

ion

deve

lop,

the

pre

cise

mut

atio

n ca

nnot

be

iden

tified

. Sin

ce t

his

band

ing

profi

le m

ay b

e ca

used

by

a m

utat

ion

likel

y no

t ass

ocia

ted

with

Am

res

ista

nce

(e.g

. c14

02t)

, re

sist

ance

to

the

drug

is in

ferr

ed u

ntil

phen

otyp

ic D

ST is

per

form

ed a

nd t

he r

esul

t use

d to

re-

eval

uate

the

tre

atm

ent r

egim

en.

SECOND LINE (SL) – LINE PROBE ASSAY (LPA) INTERPRETATION

20

FIRST AND SECOND LINE DRUGS LINE PROBE ASSAYS

Drug-resistant TB cases assessment guide based on SL-LPA

CASE 1No resistance mutations detected or inferred in any of the genomic regions included in SL-LPA.

All WT bands developed and no MUT bands developed in SL-LPA.

Con

juga

te C

ontr

olA

mpl

ifica

tion

Con

trol

TUB

gyrA

gy

rA W

T1gy

rA W

T2gy

rA W

T3gy

rA M

UT1

(A90

V)

gyrA

MU

T2 (S

91P)

gyrA

MU

T3A

(D94

A)

gyrA

MU

T3B

(D94

N/Y

)gy

rA M

UT3

C (D

94G

)gy

rA M

UT3

D (D

94H

)

gyrB

gy

rB W

Tgy

rB M

UT1

(N53

8D)

gyrB

MU

T2 (E

540V

)

rrs

rrs

WT1

rrs

WT2

rrs

MU

T1 (a

1401

g)rr

s M

UT2

(g14

84t)

eis

eis

WT1

eis

WT2

eis

WT3

eis

MU

T1 (c

-14t

)

M

Genotypic report:

Resistance not detected

Additional Diagnostic Action:

Optional: • Perform phenotypic DST for Lfx at CC (e.g. CC: 1.0 mg/L in MGIT and 7H10) and/or

for Mfx at CC and CB (e.g. CC: 0.25 mg/L in MGIT and 0.5 mg/L on 7H10; CB: 1.0 mg/Lin MGIT and 2.0 mg/L on 7H10).

• Perform phenotypic DST for SLI drug of interest.

The decision to perform these optional follow-up actions should be guided by considerations on the individual patient’s risk group for resistance (e.g. prior exposure to SL medicines, suspect treatment failure), and by the prevalence of resistance in the specific geographical setting, as these factors affect the predictive values of the test.

Clinical implications:

Start MDR-TB treatment. Review treatment regimen based on phenotypic DST results.

21

CASE 2Detection of resistance mutations associated with high-level increase in MIC for Mfx

DRUG-RESISTANT TB CASES ASSESSMENT GUIDE BASED ON SL-LPA

Con

juga

te C

ontr

olA

mpl

ifica

tion

Con

trol

TUB

gyrA

gy

rA W

T1gy

rA W

T2gy

rA W

T3gy

rA M

UT1

(A90

V)

gyrA

MU

T2 (S

91P)

gyrA

MU

T3A

(D94

A)

gyrA

MU

T3B

(D94

N/Y

)gy

rA M

UT3

C (D

94G

)gy

rA M

UT3

D (D

94H

)

gyrB

gy

rB W

Tgy

rB M

UT1

(N53

8D)

gyrB

MU

T2 (E

540V

)

rrs

rrs

WT1

rrs

WT2

rrs

MU

T1 (a

1401

g)rr

s M

UT2

(g14

84t)

eis

eis

WT1

eis

WT2

eis

WT3

eis

MU

T1 (c

-14t

)

M

If one of the following MUT probe is developed:

• gyrA MUT3C (i.e. gyrA D94G) (see picture above as example)• gyrA MUT3D (i.e. gyrA D94H)• gyrA MUT3B (i.e. gyrA D94N/Y)

Genotypic report:

Levofloxacin: Resistance detectedMoxifloxacin: Mutation associated with high-level increase in MIC for Mfx detected

Additional Diagnostic Action:

Optional: Perform phenotypic DST for SLI drug of interest.

Clinical implications:

Mfx even at a high dose cannot be considered as an effective medicine.

CASE 3Detection of mutations associated with at least low-level increase in MIC for Mfx

Con

juga

te C

ontr

olA

mpl

ifica

tion

Con

trol

TUB

gyrA

gy

rA W

T1gy

rA W

T2gy

rA W

T3gy

rA M

UT1

(A90

V)

gyrA

MU

T2 (S

91P)

gyrA

MU

T3A

(D94

A)

gyrA

MU

T3B

(D94

N/Y

)gy

rA M

UT3

C (D

94G

)gy

rA M

UT3

D (D

94H

)

gyrB

gy

rB W

Tgy

rB M

UT1

(N53

8D)

gyrB

MU

T2 (E

540V

)

rrs

rrs

WT1

rrs

WT2

rrs

MU

T1 (a

1401

g)rr

s M

UT2

(g14

84t)

eis

eis

WT1

eis

WT2

eis

WT3

eis

MU

T1 (c

-14t

)

M

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

22

If one of the following MUT probe is developed:

• gyrA MUT1 (i.e. gyrA A90V) (see picture above as example)• gyrA MUT2 (i.e. gyrA S91P)• gyrA MUT3A (i.e. gyrA D94A)• gyrB MUT1 (i.e. gyrB N538D)• gyrB MUT2 (i.e. gyrB E540D)

Genotypic report:

Levofloxacin: Resistance detected.Moxifloxacin: Mutation associated with at least low-level increase in MIC for Mfx detected.

Additional Diagnostic Action:

Recommended: Perform phenotypic DST for Mfx at CB according to Case 1. Optional: Perform phenotypic DST for SLI drug of interest.

Clinical implications:

Mfx could be used at higher dose. The regimen should be re-evaluated based on phenotypic DST results at CB.

CASE 4Precise mutation unknown, only inferred in the fluoroquinolones (FQ) regions (i.e. gyrA and gyrB)

Con

juga

te C

ontr

olA

mpl

ifica

tion

Con

trol

TUB

gyrA

gy

rA W

T1gy

rA W

T2gy

rA W

T3gy

rA M

UT1

(A90

V)

gyrA

MU

T2 (S

91P)

gyrA

MU

T3A

(D94

A)

gyrA

MU

T3B

(D94

N/Y

)gy

rA M

UT3

C (D

94G

)gy

rA M

UT3

D (D

94H

)

gyrB

gy

rB W

Tgy

rB M

UT1

(N53

8D)

gyrB

MU

T2 (E

540V

)

rrs

rrs

WT1

rrs

WT2

rrs

MU

T1 (a

1401

g)rr

s M

UT2

(g14

84t)

eis

eis

WT1

eis

WT2

eis

WT3

eis

MU

T1 (c

-14t

)

M

If one of the following WT bands is not developed:

• gyrA WT1 (i.e. gyrA WT1 probe missing) (see picture above as example),• gyrA WT2 (i.e. gyrA WT2 probe missing),• gyrA WT3 (i.e. gyrA A WT3 probe missing),• gyrB WT (i.e. gyrB WT probe missing),

and none of the MUT probe is developed in the gyrA and gyrB regions.

23

DRUG-RESISTANT TB CASES ASSESSMENT GUIDE BASED ON SL-LPA

Genotypic report:

Levofloxacin: Resistance inferred.Moxifloxacin: Mutation associated with at least low-level increase in MIC for Mfx inferred.

Additional Diagnostic Action:

Recommended: Perform phenotypic DST for Mfx at CB according to Case 1.

Optional but recommended in some settings:1 gyrA and gyrB QRDR sequencing to identify resistance mutation and exclude synonymous mutations or non-synonymous mutations not causing resistance (systematic false positive results) (interpret based on Cases 2–4 and follow respective recommendations for phenotypic DST).

If sequencing is not available, perform phenotypic DST at CC for Lfx and/or Mfx according to case 1.

Optional: Perform phenotypic DST for SLI drug of interest.

Clinical implications:

Levofloxacin is not effective. Mfx could be used at higher dose. The regimen should be re-evaluated based on phenotypic DST results at CB.

Note. These recommendations do not apply if sequencing, available before treatment initiation, identifies mutations not associated with resistance to FQ, or i f phenotypic DST shows susceptibility at CC.

1 The lack of binding of a WT probe without simultaneous binding of a mutant probe is caused by the presence of a resistance mutation (e.g. gyrA G88A). However, systematic errors are possible due to synonymous or non-synonymous mutations. Globally this is rare (<1% of isolates), but these isolates can be frequent locally. Unfortunately, it cannot be predicted in which settings these cases are frequent, which means that each laboratory has to decide whether sequencing of the QRDR region is required based on the local epidemiology. For example, the gyrA A90G mutation, which prevents binding of gyrA WT2, is frequent in the Republic of the Congo and the Democratic Republic of the Congo and a synonymous mutation codon at 96 of gyrA, which prevents binding of gyrA WT3, is frequent in Medellín (Colombia) (9). In both of these settings, sequencing would therefore be recommended.

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

24

CASE 5Detection of mutations that cause resistance to SLI.

Con

juga

te C

ontr

olA

mpl

ifica

tion

Con

trol

TUB

gyrA

gy

rA W

T1gy

rA W

T2gy

rA W

T3gy

rA M

UT1

(A90

V)

gyrA

MU

T2 (S

91P)

gyrA

MU

T3A

(D94

A)

gyrA

MU

T3B

(D94

N/Y

)gy

rA M

UT3

C (D

94G

)gy

rA M

UT3

D (D

94H

)

gyrB

gy

rB W

Tgy

rB M

UT1

(N53

8D)

gyrB

MU

T2 (E

540V

)

rrs

rrs

WT1

rrs

WT2

rrs

MU

T1 (a

1401

g)rr

s M

UT2

(g14

84t)

eis

eis

WT1

eis

WT2

eis

WT3

eis

MU

T1 (c

-14t

)

M

If one of the following MUT bands is developed:

• rrs MUT1 (i.e. rrs a1401g) (see picture above as example)• rrs MUT2 (i.e. rrs g1484t)• eis MUT1 (i.e. eis c-14t)

Genotypic report:

In case of rrs mutations only or mutations in both rrs and eis:

• Amikacin: Resistance detected• Kanamycin: Resistance detected• Capreomycin: Resistance detected

In case of eis mutation c-14t only:

• Amikacin: Resistance not detected• Kanamycin: Resistance detected• Capreomycin: Resistance not detected

Additional Diagnostic Action:

Optional: Perform phenotypic DST for FQs according to Case 1.

Clinical implications:

In case of rrs mutations only or mutations in both rrs and eis, resistance to SLI drugs is predicted.

In case of eis mutation c-14t only: amikacin is effective.

25

DRUG-RESISTANT TB CASES ASSESSMENT GUIDE BASED ON SL-LPA

CASE 6Precise mutation unknown, only inferred, in the rrs region

Con

juga

te C

ontr

olA

mpl

ifica

tion

Con

trol

TUB

gyrA

gy

rA W

T1gy

rA W

T2gy

rA W

T3gy

rA M

UT1

(A90

V)

gyrA

MU

T2 (S

91P)

gyrA

MU

T3A

(D94

A)

gyrA

MU

T3B

(D94

N/Y

)gy

rA M

UT3

C (D

94G

)gy

rA M

UT3

D (D

94H

)

gyrB

gy

rB W

Tgy

rB M

UT1

(N53

8D)

gyrB

MU

T2 (E

540V

)

rrs

rrs

WT1

rrs

WT2

rrs

MU

T1 (a

1401

g)rr

s M

UT2

(g14

84t)

eis

eis

WT1

eis

WT2

eis

WT3

eis

MU

T1 (c

-14t

)

M

If one of the following WT bands is not developed:

• rrs WT1 (rrs probe WT1 not developed) (see picture above as example),• rrs WT2 (rrs probe WT2 not developed),

and none of the MUT probes is developed in the rrs region.

Genotypic report:

In case of mutation in the rrs 1400 region:

• Amikacin: Resistance inferred1

• Kanamycin: Resistance inferred• Capreomycin: Resistance inferred

In case of mutation in the rrs 1484 region:

• Amikacin: Resistance inferred• Kanamycin: Resistance inferred• Capreomycin: Resistance inferred

Additional Diagnostic Action:

Recommended: Repeat the assay and if the result is confirmed, perform phenotypic DST for Am.

Optional: Perform sequencing to identify precise mutation. Perform phenotypic DST for FQs according to Case 1.

Clinical implications:

Kanamycin and capreomycin are likely not effective.

1 As no MUT probe in the rrs 1400 region develop, the precise mutation cannot be identified. Since this banding profile may be caused by a mutation likely not associated with Am resistance (e.g. c1402t), resistance to the drug is inferred until phenotypic DST is performed and the result used to re-evaluate the treatment regimen.

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

26

CASE 7Precise mutation unknown, only inferred, in the eis region

Con

juga

te C

ontr

olA

mpl

ifica

tion

Con

trol

TUB

gyrA

gy

rA W

T1gy

rA W

T2gy

rA W

T3gy

rA M

UT1

(A90

V)

gyrA

MU

T2 (S

91P)

gyrA

MU

T3A

(D94

A)

gyrA

MU

T3B

(D94

N/Y

)gy

rA M

UT3

C (D

94G

)gy

rA M

UT3

D (D

94H

)

gyrB

gy

rB W

Tgy

rB M

UT1

(N53

8D)

gyrB

MU

T2 (E

540V

)

rrs

rrs

WT1

rrs

WT2

rrs

MU

T1 (a

1401

g)rr

s M

UT2

(g14

84t)

eis

eis

WT1

eis

WT2

eis

WT3

eis

MU

T1 (c

-14t

)

M

If one of the following WT bands is not developed:

• eis WT1 (eis probe WT1 not developed) (e.g. eis g-37t),• eis WT2 (eis probe WT2 not developed) (e.g. eis c-12t or g-10a) (see picture above as

example),

and none of the MUT probes is developed in the eis region.

Genotypic report:

In case mutations in eis are inferred (and no additional mutations in the rrs region):

• Amikacin: Resistance not detected• Kanamycin: Resistance inferred• Capreomycin: Resistance not detected

Additional Diagnostic Action:

Optional: Perform phenotypic DST for FQs according to case 1 and for Am and Cm.

Clinical implications:

Amikacin is likely effective.

27

References

REFERENCES

1. Miotto, P et al. “A standardised method for interpreting the association betweenmutations and phenotypic drug resistance in Mycobacterium tuberculosis.” EuropeanRespiratory Journal 50.6 (2017): 1701354.

2. World Health Organization. Technical Report on critical concentrations for drugsusceptibility testing of medicines used in the treatment of drug-resistant tuberculosis.Geneva: World Health Organization;2018 (WHO/CDS/TB/2018.5). Licence: CC BY-NC-SA 3.0 IGO.http://www.who.int/tb/publications/2018/WHO_technical_report_concentrations_TB_drug_susceptibility/en/.

3. World Health Organization (2018). The use of next-generation sequencing technologies for the detection of mutations associated with drug resistance in Mycobacteriumtuberculosis complex: technical guide. Geneva: World Health Organization; 2018(WHO/CDS/TB/2018.19). Licence: CC BY-NCSA3.0 IGO.http://apps.who.int/iris/bitstream/handle/10665/274443/WHO-CDS-TB-2018.19-eng.pdf.

4. World Health Organization. (2008). Molecular line probe assay for rapid screening ofpatients at risk of multidrug-resistant tuberculosis (MDR-TB). Policy statement. WorldHealth Organization.http://www.who.int/tb/laboratory/line_probe_assays/en/.

5. Nathavitharana RR, et al. 2016. Multicenter Noninferiority Evaluation of Hain GenoTypeMTBDRplus Version 2 and Nipro NTM+MDRTB Line Probe Assays for Detection ofRifampin and Isoniazid Resistance. J Clin Microbiol 54:1624-1630.

6. World Health Organization. (2016). The use of molecular line probe assays for thedetection of resistance to isoniazid and rifampicin: policy update. World HealthOrganization. http://www.who.int/tb/publications/molecular-test-resistance/en/.

7. World Health Organization. (2016). The use of molecular line probe assays for thedetection of resistance to second-line anti-tuberculosis drugs: policy guidance. WorldHealth Organization.http://www.who.int/iris/handle/10665/246131.

8. GLI model TB diagnostic algorithms (revised June 2018). Global Laboratory Initiative.2017. http://stoptb.org/wg/gli/assets/documents/GLI_algorithms.pdf.

9. Ajileye A, et al. 2017. Some Synonymous and Nonsynonymous gyrA Mutations inMycobacterium tuberculosis Lead to Systematic False-Positive FluoroquinoloneResistance Results with the Hain GenoType MTBDRsl Assays. Antimicrob AgentsChemother 61.

10. Folkvardsen DB, et al. 2013. Can molecular methods detect 1% isoniazid resistance inMycobacterium tuberculosis? J Clin Microbiol 51:1596-9.

11. Folkvardsen DB, et al. 2013. Rifampin heteroresistance in Mycobacterium tuberculosiscultures as detected by phenotypic and genotypic drug susceptibility test methods. JClin Microbiol 51:4220-2.

LINE PROBE ASSAYS FOR DRUG-RESISTANT TUBERCULOSIS DETECTION

28

12. Nathavitharana RR, et al. 2017. Accuracy of line probe assays for the diagnosis of pulmonary and multidrug-resistant tuberculosis: a systematic review and meta-analysis. Eur Respir J 49.

13. Theron G, et al. 2016. GenoType® MTBDRsl assay for resistance to second-line anti-tuberculosis drugs. Cochrane Database Syst Rev 9:CD010705.

14. Hain Lifescience. GenoType MTBDRplus VER 2.0 Molecular Genetic Assay for Identification of the M. tuberculosis Complex and its Resistance to Rifampicin and Isoniazid from Clinical Specimens and Cultivated samples Instructions for use (June 2015).

15. Hain Lifescience. GenoType MTBDRsl VER 2.0. Molecular Genetic Assay for Identification of the M. tuberculosis Complex and its Resistance to Fluoroquinolones and Aminoglycosides/Cyclic Peptides from Sputum Specimens or Cultivated Samples. Instruction for use (June 2015).

16. Tagliani E, et al. 2015. Diagnostic Performance of the New Version (v2.0) of GenoType MTBDRsl Assay for Detection of Resistance to Fluoroquinolones and Second-Line Injectable Drugs: a Multicenter Study. J Clin Microbiol 53:2961–9.

17. WHO treatment guidelines for isoniazid-resistant tuberculosis: Supplement to the WHO treatment guidelines for drug-resistant tuberculosis. Geneva: World Health Organization; 2018. Licence: CC BY-NC-SA 3.0

http://www.who.int/tb/publications/2018/WHO_guidelines_isoniazid_resistant_TB/en/.

18. Technical report on the critical concentrations for drug susceptibility testing of medicines used in the treatment of drug-resistant tuberculosis (WHO/CDS/TB/2018.5) [Internet]. Geneva, World Health Organization; 2017. Available from:

http://apps.who.int/iris/bitstream/10665/260470/1/WHO-CDS-TB-2018.5-eng.pdf19. Technical report on the pharmacokinetics and pharmacodynamics (PK/PD) of

medicines used in the treatment of drug-resistant tuberculosis (WHO/CDS/TB/2018.6) [Internet]. Geneva, World Health Organization; 2018. Available from:

http://apps.who.int/iris/bitstream/10665/260440/1/WHO-CDS-TB-2018.6-eng.pdf20. Rigouts L, et al. 2015. Specific gyrA gene mutations predict poor treatment outcome in

MDR-TB. Journal of Antimicrobial Chemotherapy 71(2):314–23.21. Lange C, et al. 2018. Perspectives for personalized therapy for patients with multidrug-

resistant tuberculosis. Journal of internal medicine [Epub ahead of print].22. Vadwai V, et al. 2013. Can inhA mutation predict ethionamide resistance? Int J Tuberc

Lung Dis 17:129–30.23. Vilchèze C, et al. 2014. Resistance to Isoniazid and Ethionamide in Mycobacterium

tuberculosis: Genes, Mutations, and Causalities. Microbiol Spectr 2:MGM2-0014-2013.24. World Health Organization. Rapid communication: key changes to treatment of

multidrug- and rifampicin-resistant tuberculosis (MDR/RR-TB). Licence: CC BY-NC-SA 3.0.

http://www.who.int/tb/publications/2018/WHO_RapidCommunicationMDRTB.pdf.25. Ahmad N, et al. 2018. Treatment correlates of successful outcomes in pulmonary

multidrug-resistant tuberculosis: an individual patient data meta-analysis. Lancet 392:821–834.

26. Rigouts L, et al. 2013. Rifampin resistance missed in automated liquid culture system for Mycobacterium tuberculosis isolates with specific rpoB mutations. J Clin Microbiol 51:2641–5.

27. Van Deun A, et al. 2013. Rifampin drug resistance tests for tuberculosis: challenging the gold standard. J Clin Microbiol 51:2633–40.

29

28. Shah NS, et al. 2016. Clinical Impact on Tuberculosis Treatment Outcomes of Discordance Between Molecular and Growth-Based Assays for Rifampin Resistance, California 2003–2013. Open Forum Infect Dis 3:ofw150.

29. Miotto, P et al. Role of disputed mutations in the rpoB gene in the interpretation of automated liquid MGIT culture results for rifampicin susceptibility testing of Mycobacterium tuberculosis. Journal of clinical microbiology (2018): JCM-01599.

30. Seifert M, et al. 2015. Genetic mutations associated with isoniazid resistance in Mycobacterium tuberculosis: a systematic review. PLoS One 10:e0119628.

31. Kandler JL, et al. 2018. Validation of Novel Mycobacterium tuberculosis Isoniazid Resistance Mutations Not Detectable by Common Molecular Tests. Antimicrob Agents Chemother 62.

32. Andre E, et al. 2017. Consensus numbering system for the rifampicin resistance-associated rpoB gene mutations in pathogenic mycobacteria. Clin Microbiol Infect 23:167–172.

33. Camus JC, et al. 2002. Re-annotation of the genome sequence of Mycobacterium tuberculosis H37Rv. Microbiology 148:2967–73.

REFERENCES

31

ANNEX 1. FL-LPA REPORTING FORMAT – PRACTICAL EXAMPLES

Annex 1.FL-LPA Reporting format – Practical examples

Please note that the “Conclusion” section has been included for convenience but should not be part of the laboratory report.

Example 1

Drug Gene Mutation Interpretation Conclusion

Rif a rpoB H526YResistance to Rif detected.

Rifampicin is not effective.

H akatG

Mutation(s) in codon 315 region

Mutation associated with high-level increase in MIC inferred.

Isoniazid is unlikely to be effective even at high dose.

inhA t-8a

Eto/Pto inhA t-8a Resistance to Eto/Pto likely detected.

Ethionamide and prothionamide are likely not effective.

Example 2

Drug Gene Mutation Interpretation Conclusion

Rif rpoBNo mutation detected

Resistance to Rif not detected.

Rifampicin is effective.

H akatG S315T Mutation associated with

high-level increase in MIC detected.

Isoniazid is not effective even at high dose.inhA c-15t

Eto/Pto inhA c-15tResistance to Eto/Pto detected.

Ethionamide and prothionamide are not effective.

Example 3

Drug Gene Mutation Interpretation Conclusion

Rif rpoBMutation(s) in codons 516-522 (435-441)

Resistance to Rif inferred. Rifampicin is not effective.

HkatG

No mutation detected

Mutation likely associated with at least low-level increase in MIC detected.

Isoniazid at high dose is likely effective.

inhA t-8c

Eto/Pto inhA t-8cResistance to Eto/Pto likely detected.

Ethionamide and prothionamide are likely not effective.

a If more than one probe per drug that provides information the results should be reported according the following hierarchy (where the “>” sign means overrule):For H: Mutation associated with high-level increase in MIC detected > Mutation associated with high-level increase in MIC inferred > Mutation associated with at least low-level increase in MIC detected > Mutation associated with at least low-level increase in MIC inferred > Resistance not detected.For Rif: Resistance detected > Resistance inferred > Resistance not detected.

32

FIRST AND SECOND LINE DRUGS LINE PROBE ASSAYS

Annex 2.SL-LPA Reporting format – Practical examples

Please note that the “Conclusion” section has been included for convenience but should not be part of the laboratory report.

Example 1

Drug Gene Mutation Interpretation Conclusion

Lfx agyrA D94A Resistance to Lfx

detected.Levofloxacin is not effective.

Moxifloxacin could be used at higher dose. The regimen should be re-evaluated based on phenotypic DST results at CB.

gyrB No mutation

Mfx agyrA D94A

Mutation associated with at least low-level increase in MIC for Mfx detected.gyrB No mutation

Km arrs a1401g

Resistance to Km detected. Kanamycin is not effective.

Amikacin is not effective.

Capreomycin is not effective.

eis promoter

Mutation(s) in the -10 to -15 region

Am rrs a1401gResistance to Am detected.

Cm rrs a1401gResistance to Cm detected.

Example 2

Drug Gene Mutation Interpretation Conclusion

LfxgyrA A90V Resistance to Lfx

detected.Levofloxacin is not effective.

Moxifloxacin could be used at higher dose. The regimen should be re-evaluated based on phenotypic DST results at CB.

gyrB No mutation

MfxgyrA A90V Mutation associated

with at least low-level increase in MIC for Mfx detected.gyrB No mutation

Km

rrs No mutationResistance to Km inferred.

Kanamycin is not effective.

Amikacin is effective.

Capreomycin is effective.

eis promoter

Mutation(s) in the -37 region

Am rrs No mutationResistance to Am not detected.

Cm rrs No mutationResistance to Cm not detected.

33

ANNEX 2. SL-LPA REPORTING FORMAT – PRACTICAL EXAMPLES

Example 3

Drug Gene Mutation Interpretation Conclusion

LfxgyrA Mutation(s) at

codon 89–93 Resistance to Lfx inferred

Levofloxacin is not effective.

Moxifloxacin could be used at higher dose. The regimen should be re-evaluated based on phenotypic DST results at CB.

Note. These recommendations do not apply if sequencing, available before treatment initiation, identifies mutations not associated with resistance to FQ, or if phenotypic DST shows susceptibility at CC.

gyrB No mutation

Mfx

gyrA Mutation(s) at codon 89-93 Mutation associated

with at least low-level increase in MIC for Mfx inferred.

gyrB No mutation

Km

rrs No mutationResistance to Km not detected

Kanamycin is effective.

Amikacin is effective.

Capreomycin is effective.

eis promoter

Mutation(s) in the -2 region

Am rrs No mutationResistance to Am not detected

Cm rrs No mutationResistance to Cm not detected

a If more than one probe per drug that provides information the results should be reported according the following hierarchy (where the “>” sign means overrule):For Lfx, Am, Km, Cm: Resistance detected > Resistance inferred > Resistance not detected.For Mfx: Mutation associated with high-level increase in MIC detected > Mutation associated with at least low-level increase in MIC detected > Mutation associated with at least low-level increase in MIC inferred > Resistance not detected.

34

FIRST AND SECOND LINE DRUGS LINE PROBE ASSAYS

Annex 3.Specific nucleotide changes

detected by MUT probes

Please note that some of the amino acid (AA) changes identified by FL- and SL-LPA are caused by nucleotide changes that are not specifically recognized by the MUT probes. For instance, the gyrA mutation A90V is caused by two possible nucleotide changes, namely (i) gcg>gtg or (ii) gcg>gtc. However, only the former nucleotide change (gcg>gtg) will be recognized by the gyrA MUT1 probe, while the latter (gcg>gtc) will be detected only by the absence of gyrA WT2 (i.e. gyrA WT2 not detected).

MUT probe AA change Nucleotide change

rpoB MUT probes

MUT1 D516V (D435V) gac>gtc

MUT2A H526Y (H445Y) cac> tac

MUT2B H526D (H445D) cac> gac

MUT3 S531L (S450L) tcg>ttg

MUT probe AA change Nucleotide change

katG MUT probesMUT1 S315T agc>acc

MUT2 S315T agc>aca

MUT probe AA change Nucleotide change

gyrA MUT probes

MUT1 A90V gcg>gtg

MUT2 S91P tcg>ccg

MUT3A D94A gac>gcc

MUT3B D94N gac>aac

MUT3B D94Y gac>tac

MUT3C D94G gac>ggc

MUT3D D94H gac>cac

MUT probe AA change Nucleotide change

gyrB MUT probesMUT1 N538D (N499D) aac>gac

MUT2 E540V (E501V) gaa>gta

www.stoptb.org/wg/gli